TECHNOLOGY TRENDS

Technology Trends
in Audio Engineering
A report by the AES Technical Council
INTRODUCTION
Technical Committees are centers of technical expertise within the AES. Coordinated by the AES Technical Council, these
committees track trends in audio in order
to recommend to the Society papers,
workshops, tutorials, master classes, standards, projects, publications, conferences,
and awards in their fields. The Technical
Council serves the role of the CTO for the
society. Currently there are 23 such
groups of specialists within the council.
Each consists of members from diverse
backgrounds, countries, companies, and
interests. The committees strive to foster
wide-ranging points of view and
approaches to technology. Please go to:

http://www.aes.org/technical/ to learn
more about the activities of each committee and to inquire about membership.
Membership is open to all AES members
as well as those with a professional interest in each field.
Technical Committee meetings and
informal discussions held during regular
conventions serve to identify the most
current and upcoming issues in the specific technical domains concerning our
Society. The TC meetings are open to all
convention registrants. With the addition
of an internet-based Virtual Office, committee members can conduct business at
any time and from any place in the world.
One of the functions of the Technical

Council and its committees is to track

new, important research and technology
trends in audio and report them to the
Board of Governors and the Societys
membership. This information helps the
governing bodies of the AES to focus on
items of high priority. Supplying this
information puts our technical expertise
to a greater use for the Society. In the following pages you will find an edited compilation of the reports recently provided
by many of the Technical Committees
Francis Rumsey
Chair, AES Technical Council
Bob Schulein, Jrgen Herre, Michael Kelly
Vice Chairs

ARCHIVING, RESTORATION,
AND DIGITAL LIBRARIES
David Ackerman, Chair
Chris Lacinak, Vice Chair

Practical observations
Broadcast Wave File (BWF) format has
become the de facto standard for preservation of audio content within the field, as
has a digital audio resolution of 24 bit/96
kHz. Time-based metadata is also of particular interest, including time-stamped
descriptive metadata and closed captions.
Manufacturers have begun to enable preservation activities through additional
metadata capabilities and support for open
formats.
Sound for moving image is somewhat in
limbo, currently being grouped with moving image preservation for the most part.
Preservation of sound for moving image is a
current focus for future attention of this
committee. Moving image and sound
preservation graduate programs are emerg90

ing throughout the world to support those

who oversee and manage moving image and
sound archives. This is acknowledgment of
the differing skill set from traditional paper
and still image archivists.
IT and programming skills are an evergrowing need in the fulfilment of preservation. This is an emerging required understanding / skill for audio engineers.
Requirements and specifications for digital
repositories serving preservation and access
roles are currently in development.

Selected significant projects

and initiatives
The 131st AES Convention featured an
archiving track that was well attended. We
believe archiving will continue to grow as
an area of interest to AES members.

In addition, the Technical Committee on

Audio Recording and Mastering Systems
has completed a study on the persistence
and interoperability of metadata in wav
files, while Indiana University published
Meeting the challenge of media preservation; strategies and solutions.
The following standards activities
recently took place:
AES60-2011 AES Standard for audio
metadataCore audio metadata was published September 22, 2011.
AES57-2011 AES standard for audio
metadataAudio object structures for
preservation and restoration was published
September 21, 2011
AES SC-07-01 Working Group on audio
metadata was formed this October. This
group continues the work to complete AES-

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

TECHNOLOGY TRENDS
X98C, metadata for process history of audio
objects.
AES SC-03 was retired this October.
The Federal Agencies Audio Visual Digitization Working Group (digitizationguidelines.gov) is investigating audio system
evaluation tools for evaluating the performance of analog to digital converters and for
detecting interstitial errors.
The Indiana University Archives of Traditional Music (ATM) and the Archive of
World Music (AWM) at Harvard University
have received a grant from the National
Endowment for the Humanities to undertake a joint technical archiving project, a

collaborative research and development initiative with tangible end results that will
create best practices and test emerging
standards for digital preservation of archival
audio. This is known as Sound Directions.
The National Recording Preservation
Board, mandated by the National Recording
Preservation Act of 2000, is an advisory
group bringing together a number of professional organizations and expert individuals concerned with the preservation of
recorded sound. The group has published a
report from the engineers roundtable
(CLIR).
The National Digital Information Infra-

structure and Preservation Program

(NDIIPP) has the mission to develop a
national strategy to collect, archive, and
preserve the burgeoning amounts of digital
content, especially materials that are created only in digital formats, for current and
future generations.
Presto Center is a European effort to
push the limits of current technology
beyond the state of the art, bringing
together industry, research institutes, and
stakeholders to provide products and services for bringing effective automated preservation and access to Europes diverse audiovisual collections.

AUDIO FOR GAMES

Michael Kelly and Steve Martz, Chairs

Kazutaka Someya, Vice Chair
Emerging trends in audio for games are
driven by continuing advances in game
technology and the diversity of devices and
operating systems that are now considered
gaming devices. Trends are summarized
under the headings below.

and Android devices) and non-contact technology (e.g., Microsoft Kinect, PlayStation
Eye). These are able to track player position
or gestures and are beginning to find useful
applications in game-audio. 3-D video is yet
to demonstrate a new counterpart in audio.

A general move from hardware

to software processing

Spatial audio

Audio DSP is now performed in software on

CPUs or programmable DSP processors.
Even on lower-power platforms there is a
move away from dedicated audio chips and
memory although exceptions still exist.

Game platforms are diversifying

Console platforms are very dominant in
large budget titles and a lot of memory and
DSP is leveraged for audio on these platforms. Consoles remain a major driver in
game-audio trends and games often target
high-end consumer playback environments.
Portable platforms, particularly iOS and
Android devices, now also account for a
large portion of gameplay and present new
constraints, development approaches, and
creative styles. Production methodologies
for console and mobile gaming will increasingly merge as handheld devices become
more powerful. More recently, cloud gaming is demonstrating itself as a viable platform and offers new challenges including
potential latency and network delivery
issues.

Peripherals and interaction

Social gaming and new platforms offer new
ways to interact with games using handheld devices (e.g., Wii Remote, PlayStation
Move Controller), touch screens (e.g., iOS

Console games are largely geared around

5.1 and 7.1 playback or legacy formats.
Some commercial games (e.g., Race Driver:
Grid) are now making use of Ambisonics.
Portable platforms are generally targeted at
headphone playback or device speaker playback, although many tablets are equipped
with other methods such as HDMI outputs.
There is general trend toward scalability
and adaptation to the consumers configuration, particularly as the line between console and portable platforms becomes
blurred. The driver for spatial audio formats
largely comes from outside the games
industry and future conventions include
features such as height-channels to augment current multichannel setups.

Audio input
Speech input is now used in a number of
games and devices for character control or
player-to-player communication. Speech
analysis and processing is a key research
area in game-audio. Analysis of singing and
research in this area has been applied in a
number of leading console game titles.
Rhythm based games (e.g., Rock Band, Guitar Hero) make use of varying degrees of
instrument-style peripherals such as guitar
controllers, piano keyboards, virtual drum
kits; as well as motion controllers and
touch screens. New technologies, such as

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

those used in games like Rocksmith, permit

the use of real instruments as game
controllers.

DSP plugins and codecs

A move into software has made it possible
for developers to write their own DSP plugins for use in games. There has been an
increase in third-party companies providing DSP algorithms for licensing by game
developers. Solutions often involve platform-specific optimized codecs and DSP for
use in-game as well PC versions in the
form of VST plugins or similar for authoring. There has been a growth in use of
algorithms such as convolution reverb and
efforts to further R&D in improving audio
DSP for use in game. There has also a
strong trend toward returning to synthesized sound in-game; this is partially
driven by resource requirements of
portable platforms, but also by the potential flexibility of synthesized sound as new
R&D can provide improved quality for
appropriate sounds. As well as low-level
DSP, higher level systems such as intelligent or automatic mixing technologies are
being used in games like the Battlefield
series.

Tools and workflow

A number of studios now have extremely
sophisticated tools for game audio content
authoring, either developed in-house or
licensed as middleware. Tools generally
remain specific to the game domain. There
are an increasing number of attempts
though standards groups like the IASIG to
increase interoperability between linear
audio tools and game-tools.
91

TECHNOLOGY TRENDS
Education and standards
Standards activity continues in the games
industry and becomes more relevant as the
industry matures. Current standards activity includes: interoperable file formats, digital audio workstation design, and loudness

levels in games. There has been recent

growth in the number of educational institutions that offer game audio courses and
interest from academia continues to grow.
This is an important step as informed game
audio programmers are still in short supply.

The IASIG recently introduced game audio

curriculum guidelines for interested institutions. Research from academia is also
directly impacting game development and
many titles feature the results of collaboration between academia and industry.

AUDIO FOR
TELECOMMUNICATIONS
Bob Zurek, Chair
Antti Kelloniemi, Vice Chair

The trend in mobile telecommunications

has been toward moving advanced features
down in price point to feature phones and
using the more advanced mobile devices as
a personal computing and multi-media
capture and playback devices. The typical
feature phone today exhibits all of the characteristics of a top of the line device of a few
years ago with both private mode and hands
free audio, multiple microphones with
advanced noise reduction capabilities, and
Bluetooth allowing a low end feature phone
to serve as the center of a personal communications network.
Wideband audio communications has
been rolled out in many countries over
both cellular and wireless VOIP (voice over
internet protocol) doubling the audio bandwidth used in speech communications.
Multiple VOIP clients are available for
download on the major mobile operating
systems and many devices come with at
least one VOIP client preinstalled.
The last few years have shown smartphones and tablet devices becoming a
larger percentage of the total mobile
telecommunications devices. They are no
longer the niche devices of the mid to late
part of the last decade. The move to common operating systems with thousands of
applications allows the user to customize
their device in ways not possible a few years
ago. The downloadable application environments of the major mobile operating systems have allowed different users to take
the same hardware and customize it into
very diverse devices to suit their needs,
from business oriented devices, to media
and gaming devices, even as far as using
the device as a configurable piece of test
equipment.
The integration of sensing capabilities
such as accelerometers, gyroscopes, light
and infrared sensors into devices have
allowed not only manufacturers but also
creators of applications the ability to create
more natural human interfaces to the
92

device and have allowed the device to more

accurately detect the environment that it is
in. This allows the device to adapt its operation, to best function in any environment
whether the device is being used for multimedia playback, communications, or computing. Voice control of communications
devices has progressed to the point where
networked voice recognition allows the use
of natural language with larger vocabularies than previously possible on a standalone
device.
Many people have replaced several individual pieces of mobile electronics with
their portable communication device over
the last few years. Integration of high quality optics has led to the replacement of still
and video capture devices for some. Current
devices are capable of both multi-megapixel
still photography and HD video capture.
Some of the devices feature multichannel
audio capture capabilities. The combination
of GPS and network connectivity has
allowed the portable communications
devices to become personal navigation
devices with nearly continuous map
updates and real time traffic information.
The enhanced processing capabilities of
separate application processors coupled
with the over the air download of applications have led to the use of portable communications devices for office productivity,
multimedia playback, and authoring, as
well as gaming all in a single device.
Current 3G and 4G data rates allow the
mobile devices to operate with bandwidths
comparable to home-based high speed
internet. This has led to the use of wireless
devices as wi-fi hubs for a network of
devices requiring internet access such as
personal computers, gaming systems, automobiles, and televisions.
Many of the advances in handsets of a few
years ago have migrated to the edge of the
personal network allowing headphones,
headsets, and car-kits to achieve handset
levels of uplink voice quality. Consumers

can upgrade the call quality of older devices

by adding new Bluetooth headsets or car
kits that contain many of the same noise
adaptive algorithms found in much newer
devices. This includes both noise adaptive
downlink and advanced noise and echo suppression in the uplink signal.
Following the move toward using the
mobile device as a users main device for
communication, computing, and media
playback has led the creation of a number
of multimedia docks, computing docks, and
accessories for the devices. In many cases
the portable communication device can
serve as the hub for the home multimedia
system, when paired to or placed in docking
systems connected to the home audio video
system. It is not uncommon for smartphones and tablets to have HDMI output for
media playback on HDMI compatible monitors or sound systems. The creation of
Bluetooth mice and keyboards, and laptop
docks often in conjunction with HDMI
video output has allowed the user to
quickly and effortlessly transition from
using the communications device as a
portable phone to a home computer.
Software updating of not only the applications but also the operating systems
allow for the devices to grow in capability
after purchase much as personal computers
have in the past. No longer is a customer
forced to live with the limitations that a
device is shipped with for the life of the
device or service provider contract. As new
features are developed and integrated into
operating systems, as long as the hardware
still supports the new functionality, a user
of a year-old device can update to many of
the features being released in the latest
devices.
Over the next few years, the rapid growth
in capabilities of portable communication
devices tied with ever-expanding application environments will allow portable communications devices to evolve into tools
unimaginable a few short years ago.

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

TECHNOLOGY TRENDS
AUDIO FORENSICS

Jeff M. Smith, Chair

Christopher Peltier, Vice Chair
Eddy Bgh Brixen, Vice Chair

Enhancement
The enhancement of forensic audio
recordings remains the most common
task for forensic audio practitioners. The
goal of forensic audio enhancement is to
increase intelligibility of voice information or improve the signal to noise ratio
of a target signal by reducing the effects
or interferences that mask it. Many tools
are available through various software
developers with the most common being
noise reductioneither adaptive or linear. Difficulties in this area are caused by
lossy data compression common to small
digital recorders, data compression, and
bandwidth limited signals in telecommunications, and non-ideal recording environments common to surveillance and
security. One growing area of research is
the assessment of speech intelligibility
with multiple papers presented on the
topic at the AES 39th Conference on
Audio Forensics in 2010.

Authentication
The majority of audio media presented to
the forensic examiner are digital recordings on optical disc, HDD, flash memory,
and solid-state recorders. However, the
analysis of analog tape cassettes and
microcassettes is still required of examiners. In the area of forensic media authentication, digitally recorded audio files
may be subject to various kinds of manipulation that are harder to detect than
those in the analog domain. This leaves
the forensic audio examiner with new
challenges regarding the authentication
of these recordings. Many new techniques
have been developed in recent years for
use in these analyses. These techniques
continue to be published and presented
through the AES Journal and proceedings
of AES Conferences and Conventions.
Among these techniques is the analysis of
the Electric Network Frequency component (ENF) of a recording. If present, the
remains of the ENF may be compared to a
database of ENF from the same grid to
authenticate the date and time the
recording was made. In addition to automatic database comparison, it is possible
to learn several other things from ENF
analysis including whether portions of
the recording were removed, if an audio
recording was digitized multiple times,

how the recorder was powered, and more.

Recent developments in digital audio
authentication also include the Compression Level Analysis of an audio recording
to determine if an uncompressed file had
been previously subject to data compression or if the compression level present is
consistent with an authentic recording.
Also, a technique for determining the
presence of butt-splice edits has been presented. In the digital domain, as in the
analog, auditory, and spectral acoustic
analysis continues to be necessary. However, it is also clear that analysis of the
digital data that makes up a recorded
audio file including its header and file
structure must be exploited to ascertain a
digital recordings authenticity.

Speech and speaker analysis

The analysis of speech and speakers present on audio recordings is a large domain
that intersects many industries including
forensics and security. The analysis of
speakers present in recordings to ascertain identity continues to be a common
request of forensic audio examiners.
However, identifying persons in a 1:1
comparison is not supported within the
scientific community that favors recognition of persons based on extracted features relative to a background model representing a population of speakers.
Automatic systems based on cepstral coefficients, Gaussian Mixture Modeling, and
likelihood ratios employ robust and validated techniques for speaker recognition.
This quantitative approach better measures and takes into account intra- and
inter-speaker variation. When used in a
forensic environment where trained
examiners base conclusions on likelihood
ratios, this technique is valued greatly
over other qualitative analyses.
The capability of a system to process
multitudes of audio signals and sort them
based on language, topic, speakers present, and acoustic environment continues
to progress with many new advances. An
interesting area of research and its application in audio forensics is Computational Auditory Scene Analysis (CASA).
This field of audio processing is interested in developing machine systems that
perform automatic signal separation
using principles derived from the human

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

auditory systems perceptual abilities to

understand audio scenes. CASA systems
have already proven very useful as preprocessors for automatic speech recognition systems and in hearing aids. New
areas of study include their use in audio
forensics. Also, automatic speaker segmentation based on extracted spectral
features and statistical modeling can help
automated systems tasked with speech
and speaker recognition.

Other considerations
Since the fundamental aspect of forensic
audio is its application to law with the litigation process benefitting from audio
enhancement and analysis, it is important
for the practitioner working with forensic
audio to be aware of this process and the
need for proper evidence handling and
laboratory procedures. As digital audio
proliferates so to have the identification
of proper practices for imaging media,
hashing file duplicates, and recovering
and/or repairing corrupt or carved files.
Additionally, it is not only common for
forensic audio to be played in a courtroom but for typed transcripts of
recorded conversations to be prepared for
the individuals involved in a case; the
lawyers, judge(s), and/or jury. Specific to
these needs, there are developments in
addressing the inherent bias present in
the human preparation of these transcripts. Also, the forensic audio practitioner must be aware of the audio samples being presented taking into
consideration courtroom acoustics, psychoacoustics, and the hearing abilities of
these individuals.

AES activities
Numerous papers on audio forensics
appear in the Journal of AES and are presented at AES conventions each year.
Additionally, there have been three AES
conferences on audio forensics since 2005
(AES 26th, 33rd, and 39th) and the next
will be in Denver, CO in 2012. Additionally, regular workshops and tutorials
appear at AES conventions. At the AES
130th Convention in London there was a
tutorial on forensic audio enhancement,
and at the AES 131st Convention in New
York there was a workshop on forensic
audio authentication.
93

TECHNOLOGY TRENDS
AUDIO RECORDING
AND MASTERING SYSTEMS

Kimio Hamasaki, Chair

Toru Kamekawa and Andres Mayo, Vice Chairs
The growth of multichannel audio recording and production is the most remarkable
trend in the audio recording and mastering
systems. Recording and mastering using
high resolution audio technology is also a
notable trend in this area.
While 5.1 multichannel sound is widely
applied in audio recording and mastering,
audio recording and mastering using
advanced multichannel sound formats such
as 7.1, 9.1, and more channels have been
increasing. Higher sampling frequencies
such as 96 kHz and 192 kHz are also
applied in audio recording and mastering.
Most recording systems can now work at
these higher sampling rates, including in
some cases the very high rate used by DSD
(Direct Stream Digital) systems. DXD (Digital eXtreme Definition), which samples
multi-bit PCM at 352.8 kHz, is a new trend
for digital recording. A-to-D converters and
D-to-A converters for DXD are available,
and some DAWs can record and edit DXD.
The digital audio workstation (DAW) is

the principal tool for editing and mixing.

Mixing consoles are becoming an interface
for the DAW. Physical control surfaces for
mixing are sometimes not used, but instead
a virtual control surface on a PC display is
often used for recording and mastering.
DAWs use hard disks for storage, and music
recording and mastering studios also
intend to use server-based storage systems
for recordings. Network attached storage
(NAS) is widely used for audio recording.
While removable hard disks had been
widely used for audio recording and mastering, there is still no internationally standardized removable hard disk drive.
MADI (Multichannel Audio Digital Interface) has been gaining popularity in recording systems because multichannel sound
recordings need many channels compared
with 2-channel stereo recording. Stage
boxes equipped with multichannel microphone preamps and A-to-D converters are
now available with MADI output. Use of digital microphones according to the AES 42

standard is also gradually expanding in

audio recording and mixing consoles and
stage boxes equipped with AES 42 I/O are
now available. IP networking is very often
used in audio recording and mastering.
Growth of IP networking, especially considering the increase of data transfer rates, is
essential for the improvement of recording
and mastering systems.
It is common to use DSP (digital signal
processing) in recording and mastering systems. A new trend can be seen in the application of FPGAs (field programmable gate
arrays) instead of DSP, and DAWs working
on FPGA are already available. A remarkable trend in mastering systems is the
development of new plug-in audio processing software for mixing and mastering.
DAWs equipped with plug-in audio processing software are widely used for audio
production and can be purchased quite
inexpensively. The availability of such DAWs
has been changing the nature of music
productions.

AUTOMOTIVE AUDIO
Richard Stroud, Chair
Tim Nind, Vice Chair

Vehicles with built-in internet capability (via

3G, etc.) could present numerous music and
talk selections at higher quality than most
other data-reduced sources. At least one
OEM is working on personal audio to allow
people to have the same data and source
material that they have at home in the car.
Connectivity may be based on the users
mobile phone. Some OEMs are considering
using a dedicated server to control quality.
There is an interest in providing sounds
for very quiet cars such as electric vehicles.
These include engine start and engine
running sounds for inside the vehicle and
pedestrian safety sounds for outside the
vehicle.
Hard disk drives are now used in premium audio systems. These disks tend to be
smaller than state-of-art home disk drives
because of vibration requirements (40 to 80
Gbyte drives are becoming available, and
larger drives are expected soon). Disk drive
usages include navigation data and music.
Systems allow storing of many CDs from
on-board readers and music from available
94

MP3 sources via the typically included USB

connection. SSDs (solid state devices) will
replace hard drives as preferred storage
when cost permits. Premium receivers are
beginning to appear that do not include CD
players. Increasingly larger USB drives are
becoming a primary music storage medium,
along with Bluetooth-connected cell phones
with their music libraries. Download of MP3
files into vehicles by home-based RF (radio
frequency) links has been introduced.
Objective measurement is still battling
subjective listening tests as a final authority
for OEMs. SPL vs. distortion measurements
are quite good now, and directionally correct frequency response measurements are
improving. Spatial measurement capability
is being developed and evaluated.
The trend toward higher performance
audio systems is in direct conflict with
recent trends of cost and weight reduction
of components in automobiles. Increased
application of neodymium magnets may
help here. Neodymium magnet speakers,
once attractive as an affordable means of

mass reduction, have recently become

much more expensive due to neodymium
cost increases. Some reports indicate
increases of as much as eight times their
former prices. Vendors of smaller speakers
were offering neodymium magnet speakers
at prices similar to those of ferrite magnet
speakers but are struggling to do so at present. Having a strong set of specifications
will insure that sensitivity, Xmax and other
parameters are maintained in these speakers. Planar style speakers are now found in
vehicles. These are not totally flat, but have
profiles of 10 mm or less. Some examples
have shown very low sensitivity.
HD radio components are now for sale.
AM HD radio offers much higher fidelity
and FM HD offers additional program
sources. Because of the fidelity difference
on AM, rapid switching in fringe areas must
be carefully managed.
There are an increasing number of center
speakers appearing in prestige class automotive system designs. Speakers have also
appeared in the tops of front seats. Sur-

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

TECHNOLOGY TRENDS
round sound is becoming mandatory in
high-end automotive systems even when
the source is limited to two channels (so
this is implemented using upmix algorithms). Some listeners sense that some
surround systems provide limited envelopment on both stereo and much surround
source material.
There is almost universal branding of
audio systems in luxury cars, and newer
brands are emerging. The maximum number of speakers used in luxury vehicle systems seems to be leveling out at 182. Aftermarket audio now represents a very small
part of the automotive audio market. There
are still parts of the world where 5.1 and
high-level premium audio are not featured
in most vehicles audio line-ups. These sys-

tems can perhaps take advantage of inexpensive, powerful audio DSP systems to
improve performance. Rear seat audio performance may be important in China and
other countries, as some who can afford
automobiles can also afford drivers.
Voice recognition systems for telephone
and navigation functions are becoming
more sophisticated and enjoy wider application. Automatic equalization is being
offered for audio system tuning. Use of such
automatic systems can significantly speed
the tuning process but may not be ready to
completely replace tuning for on-road performance by trained listeners. Active noise
cancellation by the audio system is being
used for exhaust drone under condition of
cylinder deactivation. Active road noise bass

and/or level compensation now enjoy a

widespread market presence. Basic versions
are available in many OEM head units while
some high-end premium systems have
more sophisticated implementations. Simple systems use the speedometer signal to
apply predefined loudness curves. Others
use microphones to measure the current
cabin noise, after separating the music,
allowing more targeted equalization or
bass/level compression to be applied.
Switching audio is now commonly seen
in automotive amplifiers. Switching audio
costs are becoming comparable with older
AB amplifiers, as the heat sink requirement
is minimized. Important for electric vehicles is the low current draw under all audio
power output conditions.

CODING OF AUDIO SIGNALS

Jrgen Herre and Schuyler Quackenbush, Chairs

Overview
Audio coding has emerged as a critical
technology in numerous audio applications. In particular, it is a key component
of mobile multimedia applications in the
consumer market. Examples include
wireless audio broadcast, internet radio
and streaming music, music download,
storage and playback, mobile audio
recording, and Internet-based teleconferencing. Example platforms include digital
audio broadcast radio receivers, portable
music players, mobile phones, and personal computers. From this, a variety of
implications and trends can be discerned.
Digital distribution of content is
offered to the consumer in many formats
with varying quality / bitrate trade-off,
depending on application context. This
ranges from very compact formats (e.g.,
MPEG HE-AACv2 and MPEG USAC) for
wireless mobile distribution to perceptually transparent, scalable-to-lossless and
lossless formats for regular IP-based
distribution (e.g., MPEG AAC, HD-AAC
and ALS).
The frontiers of compression have been
pushed further, allowing carriage of fullbandwidth signals at very low bit rates to
the point where recent coding systems are
considered appropriate for some broadcasting applications, particularly relatively
expensive wireless communication channels such as satellite or cellular channels.
While such technology predominantly
makes use of parametric approaches (at
least in part) to achieve highest possible
quality at lowest bit rates, they are typi-

cally not designed to deliver transparent

audio quality (i.e., that original and
encoded/decoded audio signal cannot be
perceptually distinguished even under
most rigorous circumstances). Nevertheless, entertainment quality services over
wireless channels have been very successful. Examples of audio coding that facilitates these new markets include MPEG
HE-AACv2 and MPEG USAC.
Transform-based audio coding schemes
have been exploited to their full potential
(quality vs. bitrate). As such, new paradigms will be exploited to gain further
compression efficiency.
For broadcast-only applications where
delay is not a constraint, there is the possibility to gain further compression efficiency by exploiting large algorithmic
delays or even multi-pass algorithms in
the case of off-line audio coding.
The role of higher-level psychoacoustics and perception is becoming
increasingly important in audio coding.
Detection of auditory objects in an audio
stream, separation into auditory (as
opposed to acoustic) objects, and storage
and manipulation as auditory objects is
beginning to play a role. This will be an
important and ongoing area of research.

Hybrid and parametric coding

There is a consistent trend toward hybrid
coding techniques that employ parametric modeling to represent aspects of a signal, where the parametric coding techniques are typically motivated by aspects
of human perception. The core of most

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

successful audio coders is still largely

based on a classic filterbank based coding
paradigm, in which the quantization
noise is shaped in the time/frequency
domain to exploit (primarily) simultaneous masking in the human auditory
system. However, the recent success of
parametric extensions to the core audio
codec, in both market deployment and
standardization, illustrates this tendency
as follows.
Audio bandwidth extension technology
substitutes the explicit transmission of
the signals high-frequency part (e.g., by
sending quantized spectral coefficients)
by a parametric synthesis of high-frequency spectrum at the decoder side
based on the transmitted low frequency
part and some parametric side information that captures the most relevant
aspects of the original high frequency
spectrum. This exploits the lower perceptual acuity in the high-frequency region
of the human auditory system. An example is MPEG HE-AAC.
Parametric stereo techniques enable rendering of several output channels at very
low bit rates. Instead of a full transmission
of all channel signals, the stereo / multichannel sound image is re-synthesized at
the decoder side based on a transmitted
downmix signal and parametric side information that describes the perceptual properties (cues) of the original stereo / multichannel sound scene. Examples are MPEG
Parametric Stereo (for coding of two channels) and MPEG Surround (for full surround representation).
95

TECHNOLOGY TRENDS
Parametric coding of audio object signals provides, similarly to parametric
coding of multichannel audio, a very
compact representation of a scene consisting of several audio objects (e.g.,
music instruments, talkers, etc.). Rather
than transmitting discrete object signals,
the (downmixed) scene is transmitted,
plus parametric side information describing the properties of the individual
objects. At the decoder side, the scene
can be modified by the user according to
his/her preference, e.g., the level of a particular object can be attenuated or
boosted. A recent example for such a
technology is MPEG Spatial Audio Object
Coding (SAOC).
There has been significant progress in
the challenge of developing a truly universal coder that can deliver state of the
art performance for all kinds of input signals, including music and speech, that
has been achieved. Hybrid coders, such as
MPEG USAC (Unified Speech and Audio
Coding), have a structure combining elements from the speech and the audio
coding architectures and, over a wide
range of bit rates, perform better than
coders designed for only speech or only
audio.

Implications for technology

and consumer applications
Solid-state and hard drive-based storage
for audio has become extremely inexpensive and consumer internet connection
speeds reach into the megabits per second
range. When such resources are available,
music streaming, download, and storage
applications no longer require state of the
art audio compression. Instead, what is
occurring in the marketplace is that consumers are operating well-known perceptual coders at higher bit rates (lower com-

pression) to achieve perceptually transparent compression of music, since the

additional increment in resources required
for such operating points is relatively inexpensive. For example, consumers are opting to use MPEG Layer III (MP3) or MPEG
AAC at rates of 256 kb/s or higher to code
their music libraries for their portable
music players.
Processor speed has continued to
increase at a tremendous pace. Even with
the low-power restrictions imposed by
battery powered portable devices, the
quantity of CPU cycles potentially available for audio processing is large. Present
audio coders work in a fraction of available CPU capacity, even for multichannel
coding, and new research may be needed
to discover how to use the additional CPU
cycles and memory space. Some possibilities are improved psychoacoustic models
and sophisticated acoustic scene analysis.
Seen overall, the research in audio coding is moving to the extremes, both
toward lowest bit rates (very lossy compression using parametric coding extensions) and highest bit rates (noiseless/
lossless coding for high resolution audio
at high sampling rates/resolutions), as
well as the more complex high-level processing (scene analysis and sound field
synthesis of various sorts).
Audio coding has successfully entered
the world of telecommunication, providing low-delay high-quality codecs that
enable natural sound for teleconferencing
and video-conferencing. Such codecs
deliver full bandwidth and high quality,
not only for speech material but also for
any type of music and environmental
sound, enabling applications such as teleteaching for music. They support spatial
reproduction of sound (stereo or even
surround), which can greatly increase the

ease of communication in conferences

between several partners.
There is considerable research activity
exploring audio presentation that is more
immersive than the pervasive consumer
5.1 channel audio systems. One might
apply the label of 3-D Audio to such
explorations, since their common thread
is the use of many loudspeakers positioned around, above, and below the listener. This might range from proposed
22.2 channel systems for the consumer to
tens or hundreds of loudspeakers for
research in, e.g., wave field synthesis. Of
great interest is exploring the impact of
loudspeakers positioned above or below
the horizontal plane of the typical 5.1
channel system. When systems with a
large number of loudspeakers are considered, efficient coding of the audio speaker
signals is of paramount importance. In
addition, a flexible rendering method that
permits high-quality playback on a wide
range of conceivable consumer loudspeaker arrangements would be very
desirable. It may be that audio coding and
rendering to arbitrary loudspeaker setups
can be realized in a unified algorithm.
This will be an interesting trend to watch.
Finally, after quite some time, the digital deadlock regarding the legitimate
commercial dissemination of authorized
digital audio content has been successfully resolved, and the business models of
the music industry have embraced the
Internet. Besides a number of (mostly
legal) sources of audio (and audio-visual)
content with very limited audio quality
and free access, several successful major
distribution platforms exist now for the
electronic distribution of audio. These
download stores offer digital audio content in a variety of formats, quality levels
and protection levels.

FIBER OPTICS FOR AUDIO

Ronald G. Ajemian, Chair (USA)
Werner Bachmann, Chair (Europe)

It is clear that there are new current and

future trends in the area of fiber optics for
audio. It has been hard to ignore that more
and more companies are deploying fiber
optics in their audio/video systems. One can
witness this especially in the broadcast field
of audio/video. In the current economy
where jobs are diminishing, there is growth
for expertise with using fiber optic-based
audio/video systems. New start-up companies come to the AES Convention every year.
96

In the future, copper based systems will

be inadequate to drive the demands for
higher bit rates and bandwidth. It is clear
from just the telecommunication and
broadcast companies that everything is
becoming more integrated. Optical fiber
cables can carry multiple signals (audio,
video, clock sync/time codes, control data,
etc.) all over a single strand of fiber or two
or more if necessary. The proof is in the
application that has been proven to elimi-

nate common noise, radio-frequency interference, electromagnetic interference, and

mains hum.
Other trends include the use of fiber optic
snakes, links, networks and switchers, cables
and connectors, microphone preamplifiers,
and feeds for stage/theater live sound. Fiber
over Cat 5 or Cat 6 is an option, and fiber
used in MADI. It is likely that fiber optics will
affect every sector of audio/video and will
eventually be ubiquitous.

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

TECHNOLOGY TRENDS
HEARING AND HEARING LOSS
PREVENTION

Robert Schulein, Chair

Michael Santucci and Jan Voetmann, Vice Chairs

Introduction
The AESTC on Hearing and Hearing Loss
Prevention was established in 2005 with
five initial goals focused on informing the
membership as to important aspects of the
hearing process and issues related to hearing loss, so as to promote engineeringbased solutions to improve hearing and
reduce hearing loss. Its aims include the
following: raising AES member awareness
of the normal and abnormal functions of
the hearing process; raising AES member
awareness of the risk and consequences of
hearing loss resulting from excessive sound
exposure; coordinating and providing technical guidance for the AES-supported hearing testing and consultation programs at
U.S. and European conventions; facilitating
the maintenance and refinement of a database of audiometric test results and exposure information on AES members; forging
a cooperative union between AES members,
audio equipment manufacturers, hearing
instrument manufacturers, and the hearing
conservation community for purposes of
developing strategies, technologies, and
tools to reduce and prevent hearing loss.

Measurement and diagnosis

Current technology in the field of audiology
allows for the primary measurement of
hearing loss by means of minimum sound
pressure level audibility vs. frequency producing an audiogram record. Such a record
is used to define hearing loss in dB vs. frequency. The industry also uses measurement of speech intelligibility masked by
varying levels of speech noise. Such measurements allow individuals to compare
their speech intelligibility signal-to-noise
ratio performance to the normal population. Other tests are commonly used as well
for diagnosis as to the cause of a given hearing loss and as a basis for treatment.
Within the past ten years, new tests have
evolved for diagnosing the behavior of the
cochlea by means of acoustical stimulation
of hair cells and sensing their resulting
motion. Minute sounds produced by such
motions are referred to as otoacoustic emissions. Measurement systems developed to
detect and record such emissions work by
means of distortion product detection
resulting from two-tone stimulations as
well as hair cell transients produced from
pulse-like stimulations. Test equipment

designs for such measurements are now in

common use for screening newborn children. Additional research is being conducted directed at using such test methods
to detect early stages of hearing loss not yet
detectable by hearing-threshold measurements. The committee is currently working
to establish a cooperative relationship
between researchers in this field and AES
members, who will serve as evaluation
subjects.

Emerging treatments and technology

Currently there is no known cure for what
is referred to as sensorineural hearing loss,
in that irreparable damage has been done to
the hearing mechanism. Such loss is commonly associated with aging and prolonged
exposure to loud sounds, although it is well
established that all individuals are not
affected to the same degree. Considerable
research is ongoing with the purpose of
devising therapies leading to the activation
of cochlear stem cells in the inner ear to
regenerate new hair cells. There are, however, drug therapies being introduced in
oral form to prevent or reduce damage to
the cilia portion of hair cells in cases where
standard protection is not enough, such as
in military situations. We are beginning to
see the emergence of otoprotectant drug
therapies, now in clinical trials that show
signs of reducing temporary threshold shift
and tinnitus from short term high sound
pressure levels. New stem cell therapies are
also being developed with goals of regenerating damaged hair cells.
Hearing instruments are the only proven
method by which sensorineural hearing
loss is treated. In general the task of a hearing instrument is to use signal processing
and electroacoustical means to compress
the dynamic range of sounds in the real
world to the now limited audible dynamic
range of an impaired person. This requires
the implementation of level-dependent
compression circuits to selectively amplify
low-level sounds and power amplification
and high-performance microphone and
receiver transducers fitted into miniature
packages. Such circuitry is commonly
implemented using digital signal processing techniques powered by miniature 1-volt
zinc-air batteries.
In addition to dynamic-range improvements, hearing aids serve to improve the

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

signal-to-noise ratio of desired sounds in

the real world primarily for better speech
intelligibility in noise. Currently miniature
directional microphone systems with port
spacings in the 5-mm range are being used
to provide improvements in speech intelligibility in noise of 4 to 6 dB. Such microphones have become rather sophisticated,
in that many designs have directional adaptation circuits designed to modify polar patterns to optimize the intelligibility of
desired sounds. In addition some designs
are capable of providing different directional patterns in different frequency bands.
Furthermore, some hearing aid manufacturers have introduced products using second-order directional microphones operating above 1 kHz with some success.
In many situations traditional hearing
aid technology is not able to provide adequate improvements in speech intelligibility. Under such circumstances wireless
transmission and reception technology is
being employed to essentially place microphones closer to talkers mouths and speakers closer to listeners ears. This trend
appears to offer promise enabled by the evolution of smaller transmitter and receiver
devices and available operating-frequency
allocations. Practical devices using such
technology are now being offered for use
with cellular telephones. This is expected to
be an area of considerable technology and
product growth.

Tinnitus
Another hearing disorder, tinnitus, is commonly experienced by individuals, often as
a result of ear infections, foreign objects or
wax in the ear, and injury from loud noises.
Tinnitus can be perceived in one or both
ears or in the head. It is usually described
as a ringing, buzzing noise, or a pure tone
perception. Certain treatments for tinnitus
have been developed for excessive conditions in the form of audio masking,
however most research is directed toward
pharmaceutical solutions and prevention.
We are also seeing the emergence of electro-acoustic techniques for treating what is
commonly referred to as idiopathic tinnitus or tinnitus with no known medical
cause. About 95% of all tinnitus is considered idiopathic. These treatments involve
prescriptive sound stimuli protocols based
on the spectral content and intensity of the
97

TECHNOLOGY TRENDS
tinnitus. In Europe, psychological assistance to help individuals live with their tinnitus is a well established procedure.

Hearing loss prevention

Hearing-loss prevention has become a
major focus of this committee due to the
fact that a majority of AES members
come in contact with high level sounds as
a part of the production, creation, and
reproduction of sound. In addition, this
subject has become a major issue of consumer concern due to the increased avail-

ability of fixed and portable audio equipment capable of producing damaging

sound levels as well as live sound performance attendance. One approach to
dealing with this issue is education in the
form of communicating acceptable exposure levels and time guidelines. Such
measures are however of limited value, as
users have little practical means of gauging exposure and exposure times. This
situation represents a major need and
consequent opportunity for this committee, audio equipment manufacturers, and

the hearing and hearing-conservation

communities. In recognition of the
importance of hearing health to audio
professionals engaged in the production
and reproduction of music, this committee has scheduled its first conference
devoted to technological solutions to
hearing loss. The 47th AES International
Conference on Music Induced Hearing
Disorders will take place in Chicago, IL,
USA from June 2022, 2012. This conference will focus on new technologies for
measurement and prevention.

HIGH RESOLUTION AUDIO

Vicky Melchior and Josh Reiss, Chairs

Within the past decade, the types, distribution, and uses of audio have greatly diversified. Portables and internet sourcing have
flourished and disc sales have fallen,
although the balance between the two
varies by country. High quality audio for
formal listening has evolved simultaneously
and mirrors many of the same influences.
There is a notable broad trend toward
increasing quality in many aspects of audio,
and together with promised developments
such as cloud storage and HD streaming,
digital audio including high quality formal
listening will continue to grow and evolve.

Music sources
High resolution remains a mainstay of professional recording and archiving due to its
extended headroom, precision, and frequency capture. In the consumer marketplace, the principal current high resolution
sources are discs, especially Blu-ray, and
internet downloads. The music for these
releases reflects a range of eras and recording techniques as well as resolutions, and
may have been remastered, transcoded, or
upsampled. Thus the frequency extension
and dynamic range in some cases is less
than that of newer recordings made directly
at high resolution.
The original high resolution disc formats
have not achieved wide success although
SACDs continue to be released in small
numbers, notably in classical music. SACDcapable players continue to be available and
todays universal players may play Blu-ray
Disc (BD), DVD, SACD, and CD. Some support for Direct Stream Digital, the single bit
encoding technique behind SACD, can be
found in professional recorders, players,
and modern interfaces, but LPCM has
largely supplanted single bit techniques as
release and recording formats.
98

With the discontinuance of HD-DVD, BD

is now the higher bandwidth successor to
DVD and is well suited for high resolution
multichannel audio, both alone and in
combination with high definition (HD)
video. The format provides an optional
8 channels of 96 kHz/24 bit audio or 26
channels of 192 kHz/24 bit. The great
majority of current BDs include one or
more of these optional formats. Audio-only
discs are not yet common, but a nascent
initiative exists on the part of several small
companies to record audio-only high res
multichannel on BD without the need for a
TV monitor. Note that derivative HD discs
also exist in some regions, for example
China Blue HD in the Chinese market.
A rapid proliferation of BD-capable
devices has resulted, encompassing players,
laptops, external BD drives for PCs, PCI
cards supporting 7.1 audio with BD decoding, recorders, and home theater processors. Many, though not all, support eight
channels of high resolution audio. The
retail industry in the U.S. also reports
growing interest among ordinary consumers in BD and multichannel audio.
At least 40 websites ranging from large
aggregators to individual orchestras and
bands now exist and sell both new work and
back catalog with resolutions from
192 kHz/24 bit to 44.1 kHz/16 bit. Tracks
are principally stereo and favor classical
music, although broader genre coverage is
increasing. Websites currently sell without
copy protection. Accordingly, few releases
at the highest resolutions are available from
the major labels.
The file formats of online downloads have
coalesced around FLAC and WMA for lossless compression and WAV or AIFF for
uncompressed LPCM. The popularity of
FLAC relates to its free, open source nature

and its compatibility with most computer

operating systems. FLAC is not widely supported on mobile devices or in many lower
priced home theater (HT) systems and can
be difficult to route through an HT system
without first transcoding.

Growth of computer
and server-based audio
There is a strong trend toward adoption of
computers and file servers into all areas of
audio, especially evident in the U.S. and Far
East. For high quality audio, there are
excellent opportunities but a range of new
technical and delivery issues. The term
computer audio covers numerous configurations where the computer may act as
front end disc player or file server; may output audio via a PCI sound card, external
sound card, or motherboard ports; and may
access downloads or streamed radio and AV
from the internet. Files may be stored on
hard drives, flash, network-attached storage
(NAS), or redundant arrays with backup;
and network file servers other than a computer may act as software players.
The traditional audiophile two-channel,
music-only marketplace has embraced
computers and file servers due to the convenience of file storage and downloads. In
this market, which overlaps professional
audio, the design ethos of low distortion,
high quality engineering has spurred
manufacturer research in identifying and
eliminating technical problems associated
with computers as front end devices.
These include isolation of noisy computer
power supplies, avoidance of jittered computer clocks, RFI shielding, special attention to computer layouts by makers of
PCI sound cards, and design of digital
interfaces to avoid contaminating an
external DAC master clock with the jitter

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

TECHNOLOGY TRENDS
and noise from the PC. Examples of the
latter include asynchronous USB, PLL
chips in association with Firewire and
SPDIF, and DAC-controlled data transmission. Much ongoing effort in computer
related software aims to provide bit-accurate decoding, ripping, playback, and
transcoding.
A trend to include computer audio in
home theater is underway as well but
with a greater mix of challenges for high
quality audio. Home theater is above all a
rapidly evolving and richly diverse area of
wide price range and capability. HT components routinely support the lower resolution compressed formats streamed from
the internet and cable, and variously the
high resolution AV needed for DVD, BD,
and HDTV. Support for the file types and
resolutions typical of downloads, disc
rips, and AV from other recording or nonmovie sources may be absent. It continues to be challenging to transmit files
without invoking unwanted sample rate
conversion, unintended transcoding (e.g.,
FLAC to MP3), bit truncation, and loss of
metadata.

Improving audio quality

Transmission of high quality, high bandwidth AV signals across networks and digital interfaces is a very active arena of work.
In addition to advances in point-to-point
interfaces discussed above, development
continues on Ethernet and HDMI.
New Ethernet initiatives such as Audio
Video Bridging (AVB) promise improved
network attributes like bandwidth reservation, traffic shaping, phase synchronization
across all channels, and low latency. AVB
Ethernet is relevant to home and car systems, although the jitter performance of
DAC clocks linked to the network will need
to be assessed.
HDMI, the point-to-point connector
required for BD and HD video, has excellent
bandwidth and an Ethernet data link
(HDMI 1.4), but lacks an audio clock. HDMI
receivers must derive audio word clock
from the video pixel clock, commonly
resulting in very high jitter that affects
quality and can be audible. Some high end
receivers address the jitter and many companies are researching it but current solutions are expensive and uncommon.

Current wireless audio devices with few

exceptions are limited to 48 kHz, but components and transmission protocols are
underway that promise 96 kHz capability.
Convergence trends are strongly evident
in AV design and will certainly continue in
light of entertainment trends such as cloud
storage and streamed HD live performance.

Research
High resolution formats in general are
mature, although efforts to improve lossless compression continue. Inquiry continues into the perceptual characteristics and
audibility of sample rates above
44.1 kHz/16 bit, and of the associated filtering and data conversion processes.
Design research continues on loudspeakers, class D amps, and microphones in
support of the wide bandwidth, low distortion, wide dynamic range requirements of
high resolution. Also, surround algorithms
emphasizing enhanced spatial coding are
an especially active research area that
should be mentioned in context of high
resolution because of the improved spatial
resolution they afford.

HUMAN FACTORS
IN AUDIO SYSTEMS
Michael Hlatky, Chair
William Martens, Vice Chair
Jrn Loviscach

The Technical Committee on Human Factors in Audio Systems provides an industry

forum for questions concerning the design
of user interaction for audio applications,
the integration of audio in man-machine
interfaces (such as warning sounds, data
sonification and auditory feedback), and
the design of interfaces for musical
instruments.

Touch screens and mobile devices

With the recent advent of ubiquitous touchcontrolled computing devices, especially
the first topic has gained considerable
importance. Devices that provide touchbased on-screen manipulation such as
smartphones and tablet PCs are heavily
used to consume all things digital. Audio
software on phones or tablets, however, is
yet mostly targeted at the consumption end
of the audio commercialization chain. The
reason why we are not yet commonly seeing professional audio workstations running on a touch screens alone might be
traced back to some of the obvious shortcomings of such devices when used to work

with digital audio: Touch screens commonly lack pixel-precise navigation, parts of
the screen will be visually obstructed by the
users hand and arm when manipulating an
on-screen control, and there is relatively
little to no tactile feedback during the interaction process.
These three reasons alone make the
design of, for instance, a touch-controlled
on-screen fader quite cumbersome. While
the precision achievable by touch manipulation of an on-screen fader might be
enough to set the playback volume when
listening to MP3s on a phone, it can be by
far not enough to set parameters when
mixing music. Some manufacturers have
therefore enabled swiping gestures on
touch-controlled faders to increase precision; this does, however, take away direct
controllability, as several micro-actions
might be necessary to achieve a desired
parameter value. Furthermore, the lack of
pressure-sensitive touch screens on the
mass market renders the expressive control
of musical instruments with such devices
nearly impossible. To enable an additional

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

degree of freedom for expressive input,

some softwarefor instance Apples
GarageBand on the iPadincorporates data
from the devices accelerometer sensor
when the user plays virtual instruments
with the on-screen keyboard.
The common smartphones collection of
sensors such as the touch screen,
accelerometer, compass, GPS, microphone,
and ambient light sensor also provides a
whole new range of input capabilities that
can be leveraged in conjunction with digital
audio. There is a collection of new audio
applications that enable users to influence
the presented audio using these sensors.
Software, such as for instance Smules I
am T-Pain, RjDjs Inception App or the
Black Eyed Peas BEP360 interactive
music video, introduce a whole new level of
interactivity into the formerly lean-back
experience of listening to music. In addition, they raise the question whether music
might in future not even be generally distributed as mere audio data, but as an
application.
Interactive audio applications also pose a
99

TECHNOLOGY TRENDS
new set of problems to designers of the
common digital audio workstation (DAW).
How does a future digital audio workstation
that is targeted at producing audio for
interactive applications integrate itself well
into the development environments for the
iPhone and its siblings? Hints might be
taken from software employed to design
interactive music scores and dynamic
sounds for computer games, such as Crytecs CryEngine, or visual programming
languages such as Cycling 74 Max, or Pure
Data.

Novel game controllers

The experimental music scene has quickly
picked up off-the-shelf devices for natural
user interaction (NUI). Novel game controllers such as the Microsoft Kinect Sensor
or the older Nintendo Wii Remote have,
however, yet to arrive in the professional
audio industry. In the gaming market especially the Kinect has had a huge impact;
Microsoft reported selling more than eight
million units within the first 60 days, making this the fastest selling consumer electronics device ever.
The Kinect controller enables data
manipulation for multiple users by natural
user interaction employing multiple users
whole bodies via skeleton tracking. This
means that for instance the positions of the
users hands in three-dimensional space
can be used to control parameters, or the
software can react directly to full-body gestures. The hacking scene, such as the attendants of the industry-sponsored Music Hack
Days, embraces these devices. In the case of
the Kinect, this is fueled in particular by
the SDKs provided by the open-source com-

munity and Microsoft itself. It took only few

days after the Kinect came to market until
the first software was published by the
open-source community enabling control
of software instruments.

The cloud
Another trend to be observed at Music
Hackdays is the rise of web-based APIs
(application programming interfaces).
Whether it is finding new audio content,
processing audio, or simply listening to
music, companies such as SoundCloud, The
Echo Nest or Spotify have an API for that.
Music discovery and recommendation via
interconnected web services are topics
taken on now by Facebook and Google, and
even Pro Tools got in its tenth incarnation
equipped with a function to directly bounce
a mix to SoundCloud. Even the DAW has
moved into the cloud, with, for instance,
PowerFXs Soundation Studio or OhmForces OhmStudio.
The key benefit of these new audio production platforms are the enhanced possibilities for remote collaboration in comparison to traditional DAWs. The move to the
cloud does, however, also enable a whole
new approach to designing user interfaces
through so-called perpetual betas. As applications are running in the browser, update
cycles are frictionless, because each time
the user loads a session, a new version of
the software can be delivered. Another fact
to keep in mind is that the computing
power in the cloud is decentralized. A limit
to the number of plugins running in parallel might be a problem of the past as soon
as audio processing has moved to the cloud.
With all this computational power avail-

able in the cloud, also completely new

approaches to user experiences inside a
DAW are possible. Already today, UJAM
enables you to sing a few lines, from which
it automatically generates a complete, professional-sounding song.
A drawback of the browser-based DAWs,
however, might be that the long-learned,
known, and expected standard user interfaces provided by the operating systems
such as the default buttons or the behavior
of menus are not easily replicable inside a
web browser. With the advent of HTML5 as
a kind of operating system of a cloud-driven audio experience, such standards might
never exist again.

Modular hardware controllers

Tangible interfaces with knobs and faders
are still a big topic, and it seems that
extreme modularity is the new trend in
hardware controllers. Steinbergs CMC
Series or Euphonics Artist Series controllers can be combined in any number,
enabling the user to build a hardware controller setup for the bedroom studio or the
scoring stage, all employing the same
components.
Recent research in the HCI community
has explored the combination of touch
screen interfaces with superimposed physical controls in audio editing tasks, such as
for instance Slap Widgets by Malte Weiss
and his coworkers or Dynamic Mapping of
Physical Controls for Tabletop Groupware
by Rebecca Fiebrink and her coworkers.
These approaches seem promising to unite
the tactile controllability of physical input
devices with the configurability of a touch
screen.

MICROPHONES
AND APPLICATIONS
Eddy B. Brixen, Chair
David Josephson, Vice Chair

The microphone is an amazing device. No

other piece of audio equipment being 20 to
50 years old would be considered as a sensible choice for modern recording. However,
that is to some degree the way microphones
are regarded.

Oldies but goodies(?)

In the marketplace of today we find a lot of
old designs still being produced. A high percentage of new products brought to market
in reality are copies of aging technologies
ribbon microphones, tube microphones,
and the like. The large number of these
100

designs introduced to the market is better

explained by the opportunity of making
good business on the general assumption
that exotic looking microphones provide
exotic audio than it is by an increased level
of research in understanding and improvements of these designs.

Transducer technology
There has been no major break-through in
transducer technology during the last
years. Microelectronic mechanical systems
(MEMS) are not yet on the market for professional audio. However, in the near future

the limited signal-to-noise ratios may not

be a problem any longer.

Digital adaptation
Innovation in the field of modern microphone technology is to some degree concentrated around adaptation to the digital age.
In particular the interfacing problems are
addressed. The continued updating of the
AES42 standard is essential in this respect.
Now dedicated input/control stages for
microphones with integrated interfaces are
available. However, different widely implemented device-to-computer standards like

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

TECHNOLOGY TRENDS
USB and Firewirewhich are not specifically reserved for audiohave also been
applied in this field. Regarding the data
streams, USB3 is fully satisfactory for most
audio purposes but USB microphones are
outside standards. However they have
reached a much higher level of popularity in
semi-pro audio and home recording compared to AES42.
DSP-controlled microphones are still
developing. This includes directional pattern
control of multi-transducer units providing
steering or multichannel output for surround
recordings. These techniques are not necessarily applicable in professional audio. However, in the field of surveillance and security

recordings the applications are obvious.

Other microphone developments

More attention has been paid to the reduction of EMC problems found in an environment of increasing high frequency electromagnetic fields that are being picked-up by
microphones.
Higher order Ambisonics has taken a
central position in the search for multi-format compatibility. Other dedicated formats
for surround sound exist. However, it seems
that the 9.1/13.1 formats are forcing many
engineers to start reinventing arrays over
again. This should not be necessary.
Some technologies earlier regarded as

exotic are finding their way into practical

applications. As an example NASA has published technical briefs on a laser microphone technology that must be regarded as
a serious solution.
Battery technologyespecially for wireless microphonesis an area of great attention. Surprisingly many engineers still
prefer replaceable batteries from rechargeable. This will change.
The difficulties of getting some of the
rare earth materials for magnets may affect
the microphone selection available on the
market. In the future the effect of this
might be realized as fewer dynamic microphones or rising prices.

NETWORK AUDIO SYSTEMS

Kevin Gross, Chair

Umberto Zanghieri and Thomas Sporer, Vice Chairs
Tim Shuttleworth
This document is a compilation of contributions from numerous members of the Technical Committee on Networked Audio Systems. The committee has identified the
following important topics related to emerging audio networking technologies. Technologies that have emerged since the last
published Emerging Trends Report from the
committee in 2007 are included. To provide
structure to the report items are discussed
in order of their maturity; commercialized
technologies implemented in products available for purchase being discussed first and
embryonic concepts in early development
come up last. Other categorizations referred
to in this document are consumer market
orientation versus professional market focus,
as well as media transport methods versus
command and control protocols.

EBU N/ACIP
The European Broadcasting Union (EBU)
together with many equipment manufacturers has defined a common framework for
Audio Contribution over IP in order to
achieve interoperability between products.
The framework defines RTP as a common
protocol and media payload type formats
according to IETF definitions. SIP is used as
signaling for call setup and control, along
with SDP for the session description. The
recommendation is currently published as
document EBU Tech 3326-2008.

Audio video bridging

The Audio Video Bridging initiative is an
effort by the IEEE 802.1 task group working
within the IEEE standards organization that
brings media-ready real-time performance to

Ethernet networks. The IEEE is the organization that maintains Ethernet standards
including wired and wireless Ethernet (principally 802.3 and 802.11 respectively). AVB
adds several new services to Ethernet
switches to bring this about. The new
switches interoperate with existing Ethernet
gear but AVB-compliant media equipment
interconnected through these switches enjoy
performance currently only available from
proprietary network systems.
AVB consists of a number of interacting
standards:
802.1AS Timing and Synchronization
802.1Qat Stream Reservation Protocol
802.1Qav Forwarding and Queuing
802.1BA AVB System
IEEE 1722 Layer 2 Transport Protocol
IEEE P1722.1 Discovery, enumeration,
connection management and control
IEEE 1733 Layer 3 Transport Protocol.
AVB standardization efforts began in
earnest in late 2006. As of November 2011,
all but the P1722.1 work have been ratified
by the IEEE.

RAVENNA
A consortium of European audio companies
has announced an initiative called RAVENNA
for real-time distribution of audio and other
media content in IP-based network environments. RAVENNA uses protocols from the
IETFs RTP suite for media transport. IEEE
1588-2008 is used for clock distribution.
Performance and capacity scale with the
capabilities of the underlying network architecture. RAVENNA emphasizes data transparency, tight synchronization, low latency,
and reliability. It is aimed at applications in

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

professional environments, where networks

are planned and managed.
All protocols and mechanisms used within
RAVENNA are based on widely deployed and
established methods from the IT and audio
industry or comply with standards as defined
and maintained by international standardization organizations like IEEE, IETF, AES, and
others. RAVENNA can be viewed as a collection of recommendations on how to combine existing standards to build a media
streaming system with the designated
features.
RAVENNA is an open technology standard
without a proprietary licensing policy. The
technology is defined and specified within
the RAVENNA partner community, which is
led by ALC NetworX and supported by
numerous well-known companies from the
pro audio market.

AES X192
Audio Engineering Society Standards Committee Task Group SC-02-12-H is developing
an interoperability standard for high-performance media networking. The project has
been designated X192.
High-performance media networks support professional quality audio (16 bit,
48 kHz and higher) with low latencies (less
than 10 ms) compatible with live sound reinforcement. The level of network performance required to meet these requirements is
achievable on enterprise-scale networks but
generally not on wide-area networks or the
public internet.
The most recent generation of these
media networks use a diversity of proprietary
and standard protocols (see Table 1). Despite
101

TECHNOLOGY TRENDS
Technology

Purveyor

Date introduced Synchronization

Transport

RAVENNA

ALC NetworX

In development IEEE 1588-2008

RTP

AVB

IEEE, AVnu

Ethernet,
RTP

Q-LAN
Dante

QSC Audio
Products
Audinate

In development IEEE 1588-2008

advanced profile
(IEEE 802.1AS)
2009
IEEE 1588-2002
2006

IEEE 1588-2002

UDP

LiveWire

Telos/Axia

2004

Proprietary (native),

UDP

Table 1 Media networks

a common basis in Internet Protocol, the systems do not interoperate. This latest crop of
technologies has not yet reached a level of
maturity that precludes changes to improve
interoperability.
The X192 project endeavors to identify
the region of intersection between these
technologies and to define an interoperability standard within that region. The initiative will focus on defining how existing
protocols are used to create an interoperable system. It is believed that no new protocols need be developed to achieve this.
Developing interoperability is therefore a
relatively small investment with potentially
huge return for users, audio equipment
manufacturers, and network equipment
providers.
While the immediate X192 objective is to
define a common interoperability mode the
different technologies may use to communicate to one another, it is believed that the
mode will have the potential to eventually
become the default mode for all systems. It
will be compatible with and receive performance benefits from an AVB infrastructure. Use of the standard will allow AVB
implementations to reach beyond Ethernet
into wider area applications.
While the initial X192 target application
is audio distribution, it is assumed that the
framework developed by X192 will be substantially applicable to video and other
types of media data.

Dante
Dante is a media networking solution
developed by Audinate. In addition to providing basic synchronization and transport
protocols it provides simple plug and play
operation, PC sound card interfacing via
software or hardware, glitch free redundancy, support for AVB, and support for
routed IP networks. The first Dante product arrived in 2008 via a firmware upgrade
for the Dolby Lake Processor and since
102

then many professional audio and broadcast manufacturers have adopted Dante.
From the beginning Dante implementations have been fully IP based, using the
IEEE 1588-2002 standard for synchronization, UDP/IP for audio transport and are
designed to exploit standard gigabit Ethernet switches and VoIP-style QoS (quality of
service) technology (e.g., Diffserv). Dante
is evolving with new networking standards.
Audinate has produced versions of Dante
that use the new Ethernet Audio Video
Bridging (AVB) protocols, including IEEE
802.1AS for synchronization and RTP
transport protocols. It is committed to supporting both IEEE 1733 and IEEE 1722.
Existing Dante hardware devices can be
firmware upgraded as Dante evolves, providing a migration path from existing
equipment to new AVB capable Ethernet
equipment.
Recent developments include announced
support for routing audio signals between
IP subnets and the demonstration of low
latency video. Audinate is a member of the
AVnu Alliance and the AES X192 working
group.

Q-LAN
Q-LAN is a third-generation networked
media distribution technology providing
high quality, low latency, and ample scalability aimed primarily at commercial and
professional audio systems. Q-LAN operates over gigabit and higher rate IP networks. Q-LAN is a central component of
QSCs Q-Sys integrated system platform.
Q-Sys was introduced by QSC Audio Products in June 2009. Q-LAN carries up to 512
channels of uncompressed digital audio in
floating point format with a latency of 1
millisecond.

WAN based telematic/distributed

performance and postproduction
Telematic or distributed performances are

events in which musicians perform

together synchronously over wide area networks, often separated by thousands of
miles. The main technical challenge associated with these events is maintaining sufficiently low latencies for the musicians to
be able to play together, given the distances involved. Emerging enabling technologies such as the low latency codecs
CELT, which stands for Constrained
Energy Lapped Transform, Opus, a merging of CELT and SILK (a Skype codec) as
well as ULD, which refers to Ultra-LowDelay allow streaming over DSL or cable
end point connections rather than highbandwidth managed networks, such as
Internet2, which are recently more commonly used.
Another wide area networked emerging
use case is streaming audio for cinema
postproduction, in which studios and postproduction facilities are connected with
one another via high-bandwidth managed
fiber networks. This allows studios to see
and hear the latest version of a film in
postproduction without having to physically move the assets to the studio or use a
file-transfer system. Real-time streaming of
uncompressed audio and video also allows
greater collaboration between directors
and postproduction facilities and between
different departments in the postproduction process.
Networked postproduction uses two
methods (at present) for streaming audio:
when audio is streamed independently of
video, hardware Layer 3 uncompressed
audio-over-IP devices are used. When audio
is streamed along with video, it is embedded in an HD-SDI video stream, and the
stream is networked using a video codec.
The former case is primarily used for audio
postproduction, in which the audio engineers are mixing to a poor-quality version
of the video; the video is then sourced
locally at all locations, and the audio
synced to it. Control information is
streamed between all nodes using high-definition KVM-over-IP devices, along with
MIDI-based control surfaces connected via
Apples MIDI Network Setup. KVM over IP
is a server management technology.
(Streaming of Ethernet-based control surfaces is forthcoming.) Video-conferencing
to allow collaboration uses either H.323
devices or the same codec used to stream
content video. Clock synchronization
between nodes can be accomplished either
with the hardware audio-over-IP devices,
which usually stream clock information, or
with GPS-based sync generators at each
node.

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

TECHNOLOGY TRENDS
XFN command and control protocol
XFN is an IP-based peer to peer audio network control protocol, in which any device
on the network can send or receive connection management, control, and monitoring
messages. The size and capability of devices
on the network will vary. Some devices will
be large, and will incorporate extensive
functionality, while other devices will be
small with limited functionality. The XFN
protocol is undergoing standardization
within the AES, and AES project X170 has
been assigned to structure the standardization process. A draft standards document
has been written and presented to the SC02-12 working group for approval.

Home broadband audio-over-IP

and home wireless LAN
Home broadband connections are increasing in speed, up to a typical rate, worldwide, of about 2 Mbps. This is sufficient for
streaming audio services to produce a good
performance, mostly using 256 kbps WMA
or AAC, which yields pretty good quality at
a low bit rate.
Use of wireless LANs in the homes,
mostly WiFi, with some proprietary systems is increasing. IEEE802.11g routers
and devices are realizing faster throughput
rates, while IEEE802.11n achieves
improved range, improved QoS, and speeds
that exceed the needs of low bit rate compressed audio streaming. Two eco-systems
co-exist at the moment. The first is the
Digital Living Network Alliance (DLNA),
which focusses on interoperability between
devices using UPnP (Universal Plug and
Play) as the underlying connectivity technology. DLNA is becoming available in
more and more devices, such as PC servers
and players, digital televisions with network connectivity, network attached storage (NAS) drives, and other consumer
devices. The second eco-system is Apple
AirPlay, which allows iTunes running on a
PC or MAC to stream audio to multiple
playback devices. AirPlay also supports
streaming directly from an iOS device
(iPhone, iTouch, iPad) over WiFi to a networked audio playback device. Both ecosystems are driving the rapid acceptance of
audio networking in the home.
Cloud computing, in particular cloud
storage of audio content, is another emerging trend. The increasing popularity of premium audio services, for example Rhapsody, Pandora, Last.fm, and Napster, are
driving a trend away from users the need to
keep a copy of their favorite music in the
home or on a portable device. Connection
to the internet allows real time access to a

large variety of content. Apple is also driving this trend with iCloud, released with
iOS5. Consumer devices are becoming
more complicated and connecting devices
to the network has been difficult for users,
resulting in many calls to tech support.
The good news is that devices are becoming easier to set up. The WiFi Alliance has
created an easy setup method call WiFi
Protected Setup (WPS). This makes attaching a new device onto the home network as
easy as pressing a button or entering a simple numeric code.
Another trend driven by the adoption of
home wireless LAN technologies is in the
user interface (UI) of networked audio
devices. More and more audio products are
using the iPhone or iPad as the primary
method of device control, via the home
WiFi network. Some commentators are
even announcing the death of the infrared
remote control. Consumer Audio/Video
Receiver manufacturers such as Denon and
Pioneer offer free iPhone/iPad apps which
allow complete, and intuitive control of
their devices. This leads to another emerging trend, that of the display-less networked audio player. Once the player can
be conveniently controlled from your
smartphone, it may not be necessary for
the device to continue to include an expensive display and user controls. Display-less
high end audio players are already selling
well (for example B&W Zeppelin Air). Such
display-less networked audio players will
become ubiquitous and be available for
under $100.

Open Control Architecture Alliance

(OCA)
The Open Control Architecture Alliance has
been formed by a group of professional
audio companies who are working in different product markets and represent a
diverse cross section of vertical market
positions and application use-cases. Each of
the companies realized that relying solely
on proprietary solutions for media networking system controls made interoperability
with other manufacturers equipment or
across application domains difficult.
The member companies agreed that an
open standardized control architecture was
not only possible, but should be created and
made available as an open, public standard
that could be available to any participant in
the audio market in order to facilitate an
improved environment for the entire AV
industry. It is the stated mission of the OCA
Alliance to secure the standardization of
the Open Control Architecture (OCA), as a
media networking system control standard

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

for professional applications. OCA in its

current form is a Layer 3 protocol that has
been created by Bosch Communications
based around the earlier (abandoned) command and control protocol AES-24. The
Alliance has been formed to complete the
technical definition of OCA, then to transfer
its development to an accredited public
standards organization.
The founding group of OCA members is
proceeding to complete the OCA specification and prepare it for transfer to a public
standards organization without inviting
new active members but welcomes any
interested parties to join as an Observer
Member.

International Telecommunications
Union: Future Networks
ITU-T Q21/13, Study Group SG13 is looking at Future Networks, which are
expected to be deployed during 20152020.
So far an objectives and design goals document has been published (Y.3001), and the
study group is working on virtualization
and energy saving (soon to be published as
Y.3011 and Y.3021 respectively) and on
identifiers. These deliberations are at a very
early stage and a clear direction is not yet
apparent. The underlying technology could
be a clean slate design, or it could be a
small increment to NGN (Next Generation
Network, which is based on IPv6).

IEC/ISO: Future Network

ISO/IEC JTC1/SC6/WG7 is also working on
Future Network, and also expects deployment during 20152020. Their system will
be a clean slate design with a control
protocol that is separate from the packet
forwarding. It will support multiple networking technologies, both legacy technologies such as IPv4 and IPv6 and also
new technologies able to provide a service
suitable for the most demanding live audio
applications.
It will carry two kinds of data, synchronous and asynchronous. For synchronous data there is a set-up process (part of
the control protocol) during which
resources can be reserved. The application
requests QoS parameters (delay, throughput, etc.) appropriate to the data to be sent,
and the network reports the service the
underlying technology is able to provide.
Asynchronous data can use a similar setup process, or can be routed in a similar
way to Internet Protocol. Thus it will also
be efficient at carrying protocols such as
TCP and will interoperate with IP networks.
This provides a migration path from current systems.
103

TECHNOLOGY TRENDS
SEMANTIC AUDIO ANALYSIS
Mark Sandler, Chair
Dan Ellis and Jay LeBoeuf, Vice Chairs
Gyorgy Fazekas

The scope of Semantic Audio Analysis has

undergone a dramatic expansion over the
past few years. As seen by many researchers
and practitioners, the area is now best
defined as the confluence of a multitude of
technologies. These include digital signal
processing tools that enable the extraction
of characteristic features from audio,
machine learning tools that connect raw
feature data with high-level semantic representations of musical content, information
management tools that allow us to effectively link and organize this information,
and knowledge representation tools facilitating the use of automatic data aggregation and high-level inferences, thus the formulation of complex queries involving
unique features of content, as well as social
metadata about musical recordings. Webbased applications that allow us to pose
queries like find me upbeat and catchy
songs between 130140 bpm, performed by
artists collaborating in the London-Shoreditch area, and sort them by musical key
are now imminent. The TC is concerned
with overseeing and coordinating developments, disseminating knowledge, and promoting novel interdisciplinary tools and
applications in the light of emerging trends
in Semantic Audio. The most important of
these novel trends and applications include
the following.

Multi-modality and the use

of contextual information
The process by which human beings understand music, and assign high-level semantic
descriptions to musical events depends on a
variety of information sources; precepts
from different senses, memory, and expectations. As recently demonstrated during
the first and highly successful AES conference on Semantic Audio Analysis in Ilmenau, Germany, researchers have started to

recognize that semantic labelling of music

relies on a number of different inputs, and
started to develop techniques that take contextual information into account. This may
be defined as a piece of complementary data
that improves the results, but not in itself
sufficient in a particular information
extraction or audio processing task. Examples of new methods include informed
source separation, which works by encoding information about the mixing process
into the stereo signal, and enhance signal
separation by using this data in the decoder,
and informed music transcription, which
takes prior information about the instruments into account. We can also observe
the increasing use of studio stems, taking
advantage of the multitrack format and the
use of multiple modalities, the simultaneous analysis of audio, video, text, and other
sources of information.

Ontologies and linked data

The heterogeneity and open-ended nature
of musical data is often the culprit of
developing complex systems that use many
sources of information. Recent developments in other disciplines, namely WebScience and the Semantic Web help us in
developing methods for associating musical data with explicit meaning in a
machine-processable way. Technologies
such as the Resource Description Framework (RDF) and Semantic Web ontologies
enable us to represent information and
knowledge in a uniform interoperable
framework, and lead to intelligent music
processing tools of the future. Semantic
Web ontologies such as the Music Ontology also provide the back-bone of the
Linked Data, which eases linking and
aggregation over disparate resources, containing increasing amounts of editorial
and social data about music.

Educational games
New advances in semantic audio technologies enable the creation of interactive educational games for music learners. It is now
possible to analyze the sound played on real
instruments and thus avoid the need for
using MIDI controllers, extract symbolic
information such as chords or note names,
and align this information with musical
scores in real-time. Applications like
Song2See demonstrate how semantic audio
technologies may help to create content for
music learners by using automated transcription, keep the user in the loop by allowing the correction of transcription errors,
use the content to ease the learning process
with fingering suggestions for each instrument, and provide real-time feedback about
the quality of playing by means of sound
analysis. The appearance of web-based platforms for content and metadata sharing,
and advances in semantic analysis and recommendation technologies also provide for
creating novel applications for music education. There is a growing trend in using community created web content, including lead
sheets and chord charts, and to analyze
YouTube videos to enhance machine analyses, or to create interactive games that are
not limited to expert generated content. The
use of the web thus provides an advantage
over games like Rock Band or Guitar Hero.

Intelligent music production tools

Finally, there is a recent increase in adapting semantic audio technologies in music
production. Examples of these applications
include navigating sound effect libraries by
using similarity defined by proximity in a
characteristic feature space, using automatic audio-to-score alignment in audio
editing, and developing intelligent audio
effects and automatic mixing techniques
that rely on semantic audio analysis.

SIGNAL PROCESSING
FOR AUDIO
Christoph Musialik, Chair
James Johnston, Vice Chair

Signal processing applications in audio engineering have grown enormously in recent

years. This trend is particularly evident in digital signal processing (DSP) systems due to
104

performance improvements in solid-state

memory, disk drives, and microprocessor
devices. The growth in audio signal processing applications leads to several observations.

Observations
First, DSP has emerged as a technical
mainstay in the audio engineering field.
Paper submissions on DSP are now among

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

TECHNOLOGY TRENDS
the most popular topics at AES Conventions, while just some years ago DSP sessions were rare at AES. DSP is also a key
field for other professional conferences,
including those sponsored by IEEE and
ASA.
Second, the consumer and professional
marketplaces continue to show growth in
signal processing applications, such as
increasing number of discrete audio channels, increasing audio quality per channel
(both word width and sampling frequency,
and increasing quality of building block
electronics, such as sampling rate
converters, ADCs and DACs, due to continuously growing availability of consumer-ready DSP hardware
Third, there is growing interest in intelligent signal processing for music information retrieval (MIR), like tune query by
humming, automatically generating
playlists to mimic user preferences, or
searching large databases with semantic
queries such as style, genre, and aesthetic
similarity.
Fourth, there are emerging algorithmic
methods designed to deliver an optimal
listening experience for the particular
audio reproduction system chosen by the
listener. These methods include transcoding and up-converting of audio material to
take advantage of the available playback
channels, numerical precision, frequency
range, and spatial distribution of the playback system. Other user benefits may
include level matching for programs with
differing loudness, frequency filtering to
match loudspeaker capabilities, room cor-

rection, and delay methods to synchronize

wavefront arrival times at a particular listening position.
In professional sound reinforcement,
loudspeakers with steerable radiation patterns can provide a practical solution for
difficult rooms. Professional live audio
applications often demand low-latency systems, which remain challenging for DSP
because many algorithms and processors
are optimized for block processing instead
of sample-by-sample, and thus introduce
more latency.
Algorithmic developments will continue
to occur in many other areas of audio
engineering, including music synthesis,
processing and effect algorithms, intelligent noise reduction in cars, as well as
enhancement and restoration for archiving and audio forensic purposes. Also,
improved algorithms for intelligent ambient noise reduction echo cancellation,
acoustical feedback suppression, and steerable microphone arrays are expected in
the audio teleconferencing field.
Fifth, switching amplifiers (like Class D)
continue to replace traditional analog
amplifiers in both low power and high
power applications. Yet even with Class D
systems, the design trends for load independence and lowest distortion often
include significant analog signal processing elements and negative feedback features. Due to advances in AD/DA converter
technology, future quality improvements
will require the increasingly scarce pool of
skilled analog engineers to design input
stages like microphone preamps, analog

clock recovery circuits, and output amplifiers that match the specifications of the
digital components.

Implications for technology

All of the trends show a demand for ever
greater computational power, memory
capacity, word length, and more sophisticated signal processing algorithms. Nevertheless, the demand for greater processing
capability will be constrained by the need
to minimize power consumption, since a
continuously growing part of audio signal
processing will be done in small, portable,
wireless, battery-powered devices. On the
other hand, due to the increasing capabilities of standard microprocessors, contemporary personal computers are now fast
enough to handle a large proportion of
the standard studio processing algorithms. Advanced algorithms still exceed
the capabilities of traditional processors,
so we see a trend in the design of future
processors to incorporate highly parallel
architectures and the compiler tools necessary to exploit these capabilities using
high-level programming schemes. Due to
the relentless price pressure in the consumer industry, processors with limited
resolution will still challenge algorithm
developers to look for innovative solutions
in order to achieve the best price-performance ratio.
The Committee is considering forging
contacts with digital signal processor
manufacturers to convey to them the
needs, experiences, and recommendations
from the audio community.

SPATIAL AUDIO

James Johnston and Sascha Spors, Chairs

Loudspeaker layouts
Nowadays surround sound is available in
many households, where the 5.1 layout is the
most deployed loudspeaker configuration.
The production chain from recording, coding, transmission to reproduction of surround sound for cinema is also well established. So far, the consumer market for
surround sound has mainly been driven by
movie titles; audio-only content is still quite
rare. As successor of the 5.1 layout, various
layouts with more loudspeakers arranged in a
plane have been proposed, for instance the
7.1 layout. None of them had the commercial
success of the 5.1 layout. Layouts that allow
for the reproduction of height seem to be the
next natural step in the evolution of surround sound. A number of proposed layouts

that include elevated loudspeakers, for

instance 9.1, 10.2 or 22.2, are becoming
ready for the market. It remains to be seen
whether the market accepts the increased
number of loudspeakers that have to be
installed at home. From a perceptual point of
view, including height cues into the reproduction has a clear benefit. However, the
optimal speaker layout is subject to a vital
discussion within the community.
Novel recording, production, and coding
techniques have to be developed and established for the new layouts including height.
Upmixing algorithms for content that has
been produced for systems not featuring
height, for instance from stereo or 5.1 surround, to the novel formats including height
are being developed. A number of proposals

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

exist, however there is still a lot of room for

new developments that can be foreseen to
show up in the future. In addition, new delivery methods that provide specific information related to height and distance, as well as
horizontal angle are being reported on at
AES conventions.

3-D
With the increased spread of 3-D video in cinema and home cinema, new requirements
must be met by spatial audio reproduction.
While 3-D video adds depth to the image, this
is not a straightforward task with stereophonic techniques. This holds especially for
sound sources closer to the listener than the
loudspeakers. Future spatial audio techniques have to provide solutions to the chal105

TECHNOLOGY TRENDS
lenges imposed by 3-D video. First concepts
have been presented.

Sound field synthesis

As alternatives to the traditional stereophonic
systems, sound field synthesis techniques
such as Wave Field Synthesis (WFS) and
higher-order Ambisonics (HOA) are being
deployed more and more. Sound field synthesis approaches are based on the principle of
physically synthesizing a sound field. In the
past two decades around 100 WFS systems
have been installed worldwide with each up
to 832 channels. Large scale Ambisonics systems are currently not so widespread, but it
seems that such systems will show up in the
near future. A vital research community
exists for both WFS and HOA that investigates various aspects and combinations of
both approaches.

Psychoacoustic motivation
Upcoming trends in spatial audio reproduction besides traditional stereophony are
multichannel reproduction systems that
are psychoacoustically motivated. Several
techniques have been developed on the
basis of WFS that aim at spatial reproduction with almost arbitrary layouts using a
decreased number of loudspeakers compared to traditional WFS. Such approaches
are already commercially available. Multichannel time-frequency approaches use
techniques from short-term signal analysis
to analyze and synthesize sound fields.
Directional Audio Coding (DirAC) and Binaural Cue Coding (BCC) are representatives
of the latter techniques. Time-frequency
processing seems to be a promising concept since its basic idea is related to the
analysis of sound fields performed by the
human auditory system.
The psychoacoustic mechanisms underlying the perception of synthetic sound fields
have been investigated in quite some detail.
However there are still plenty of open issues

in the field that should be researched in the

future. Peer-reviewed, published perceptual
research based on established psychoacoustic
methodologies will definitely bring the community forward in this aspect.

Production and mixing techniques

So far, the traditional production and mixing
techniques used in stereophony are channelbased. The different tracks are mixed in the
studio for a particular target layout and then
transmitted/stored. This process relies on the
assumption that the setup at the receiver
matches the setup used during production.
Object-oriented audio, as an alternative
approach, is based on the concept that each
track or group of tracks forms an audio
object. The signal(s) of the object together
with side information (position, room,
effects) is then transmitted to the receiver.
Here the loudspeakers signals are generated
by a suitable rendering algorithm, on the
basis of the transmitted side information. A
major benefit of the object-oriented approach
lies in the fact that it is almost independent
from the setup used by the receiver. It seems
that in the future a combination of both
approaches might be promising to cope for
the needs of the producers on the one side
and to allow setup independent mixing/production on the other side. Currently several
different formats for the transmission of the
content/side information have been proposed, however, none have yet been commercially adopted in any significant fashion.

Headphone listening
Although spatial audio is routinely used by
the gaming industry, advanced techniques
with better quality and realism can be
expected with further increases in processing
power. This holds especially for mobile
devices, where spatial audio is currently
rarely deployed. Due to the general shift
toward mobile devices spatial audio will also
be finding its way into the mobile world. As a

consequence, an increasing number of individuals use headphones when listening to

audio. In such scenarios the use of reproduction techniques based on head-related transfer functions provides truly three-dimensional spatial audio by relatively simple
technical means. As a consequence binaural
audio is expected to play a more prominent
role in the future. As an alternative to headphone listening, near-field loudspeaker playback with cross talk cancellation may be
used.

Diverse applications
Besides its traditional application fields, cinema and home cinema, spatial audio is
increasingly being deployed in other areas.
For instance, in teleconferencing systems,
cars, and as auditory system for advanced
human-machine interfaces. Here the use of
spatial audio is expected to provide a clear
benefit in terms of naturalness and transport
of additional information. Another important
area of application is virtual concert hall and
stage acoustics using active architecture systems where spatial audio enhances the environment with which musicians and audiences interact during performance. Modern
multichannel systems offer adjustability of
acoustics and high sound quality suitable for
live performance and recording.

Network standards
With respect to cabling, coping with the ever
increasing number of loudspeakers, the new
IEEE standards for Audio Video Bridging
(AVB) seems promising. The standards are
designed for the fully synchronized transmission of a high number of output/input channels via Ethernet. The standards are developed and currently supported by all major
players in the field and devices are expected
to be available in the near future. Such standards that work with intelligent processing to
detect the listening setup are expected to be
proposed soon.

TRANSMISSION AND BROADCASTING

Kimio Hamasaki and Stephen Lyman, Chairs
Lars Jonsson and Neville Thiele, Vice Chairs

The growth of digital broadcasting is the

most remarkable trend in this field. Digital
terrestrial TV and radio broadcasting have
been launched in several countries using the
technology standards listed below. Analog
broadcasting has ceased in some countries.
The World Wide Web has become a more
common, alternate source of streamed or
downloadable programming.
106

Digital terrestrial TV broadcasting

In Europe DVB-T2 has been deployed in
several countries for HD services. ATSC is
used in USA, Canada, and Korea, while
ISDB-T is employed in Japan and Brazil.

Digital terrestrial radio broadcasting

In Europe and Asia DAB+ is state of the art
in the DAB Eureka 147 family. HD-Radio or

IBOC fulfills this role in the USA and

Canada, with ISDB-SB in Japan. Large
broadcasting organizations in Europe and
Asia, and major countries like India and
Russia with large potential audiences, are
committed to the introduction of DRM
(Digital Radio Mondiale) services and it is
to be expected that this will open the market for low-cost receivers.

J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

TECHNOLOGY TRENDS
Digital terrestrial TV broadcasting
for mobile receivers
DVB-T2 Lite (Europe) is still under development, while ISDB-T is used in Japan. DMB
is employed in Korea and there have been a
few trials in Europe.
In the U.S., the Advanced Television Systems Committee (ATSC) has published final
reports of two critical industry planning
committees that have been investigating
likely methods of enhancing broadcast TV
with next-generation video compression,
transmission and Internet Protocol technologies and developing scenarios for the transmission of three-dimensional (3-D) programs
via local broadcast TV stations. The final
reports of the ATSC Planning Teams on 3-DTV (PT-1) and on ATSC 3.0 Next Generation
Broadcast Television (PT-2) are available now
for free download from the ATSC web site.

Loudness
Loudness and True Peak measurements are
replacing the conventional VU/PPM methods
of controlling program levels. This has
largely eliminated significant differences in
the loudness of different programs (and
advertisements) and the need for listeners to
keep adjusting their volume controls. Supporting international standards and operating practices have been published by several
organizations such as ITU-R, EBU and ATSC
listed below. More and more broadcasters
now apply these standards in their program
production and transmission chains.
ITU-R: BS.1770: Algorithms to measure
audio programme loudness and true-peak
audio level; BS.1771: Requirements for
loudness and true-peak indicating meters.
The following five documents provide the

core of the EBU Loudness work:

EBU R128 Loudness Recommendation
EBU Tech 3341 Metering specification
EBU Tech 3342 Loudness Range
descriptor
EBU Tech 3343 Production Guidelines
EBU Tech 3344 Distribution Guidelines
ATSC: A/85 Techniques for Establishing
and Maintaining Audio Loudness for Digital
Television.

Lip sync
The lip-sync issue remains unsolved, but is
being discussed in digital broadcasting
groups. Some international standards
development organizations such as IEC and
SMPTE are discussing new standards for
measuring the time differences between
audio and video.

Benefits of digital broadcasting

The introduction of digital broadcasting has
introduced such benefits as High Definition
TV (1080i, 720p). Due to the current availability of 5.1 surround sound in digital
broadcasting, surround sound is an important trend in TV broadcasting. 5.1 surround
sound is evolving with future extensions
involving additional channels. Along with 3D TV, several broadcasters are experimenting
with 3-D audio (for instance 22.2, Ambisonic,
wave-field synthesis, directional audio coding). Data broadcasting now includes additional information related to a program.

Digital broadcasting themes

at AES conventions
Recent AES conventions have discussed the
following digital broadcasting issues: strategies for the expansion of digital broadcasting;
audio coding quality issues for digital broadcasting and transmission; the role and
importance of audio in an era of digital multimedia broadcasting; new program-production schemes for audio in digital multimedia
broadcasting; the future of radio including
multicasting over the web and surround.

Subscribe to the AES E-Library

h ttp://www.aes.org/e-li b/subscribe /

G ai n i m m e di a t e a c ce s s t o
o v e r 1 3 , 0 0 0 f u l l y se a r c h a b l e
P D F f i l e s d o cu m e n t i n g a u di o
r e s e ar ch f ro m 1 9 5 3 t o t h e
pre sen t day. The E-librar y
i n cl u d e s e v e r y A E S p ap e r
p u b l i s h e d at a c o n v e n t i on ,
c o n f e re n c e o r i n t h e J o u r n al

Internet streaming
The use of new methods for the distribution
of signals to the home via the Internet with
streaming services is an increasing trend.
Web radio and IPTV are now getting audience figures that in a number of years from
now will be closing in on the traditional systems. Distribution technologies with rapid
growth in many countries are: ADSL/VDSL
over copper or fiber, combined with WiFi in
homes; WIMAX and 3G/UMTS; 4G and wi-fi
hot spots for distribution to handheld
devices.
J. Audio Eng. Soc., Vol. 60, No. 1/2, 2012 January/February

Individual annual subscription

$255 non-members
$145 members
Institutional annual subscription
$1750 per year
107